Motor selection guidelines for precision motion control
Several types, makes, and models of motors are available. Usually, you can group them. AC servomotors are appropriate for some applications. So too, are dc servomotors. Step motors are dc motors of a sort. To avoid confusion and to simplify the point of this article, I will divide motor types into:
• Step motors, associated with openloop motion control.
• Servomotors, including ac and dc closed-loop servos.
Some step motors serve in closed-loop systems, though not so frequently. In general, though it is possible to build a closed-loop step motor system, the cost of feedback devices and other ancillary components makes the step motor-servomotor system cost comparison approximately a draw. “Step motors” as used in this article refers to step motors in openloop mode unless stated otherwise.
Choosing a motor type
In deciding whether to use an openloop step motor or closed-loop servomotor, five simple questions about the application will initially reveal the most appropriate motor. Once you decide that, you can select from the torque-vs.-speed graphs included in suppliers’ technical data. Some additional selection criteria are presented here. They are useful in optimizing the selection.
Ask these five questions:
Is the load variable or constant? For example, office automation products such as copiers, facsimile machines, and printers, have constant-torque loads. The load is usually paper. It is predictable, and it does not vary through a move cycle. Here, you can use step motors. Conversely, factory automation applications typically have unpredictable, variabletorque loads, and they may require feedback operation. Servomotors, because of their closed-loop configuration and overload capacity, can intermittently absorb variable-torque loads during a move or process and recover to the desired position. Step motors have no tolerance for overload and must be sized for the absolute worst case. Recommendation: If you don’t want to lose position, pick a step motor with 30% more holding torque capability than the maximum torque requirement at the motor shaft.
Will the motor operate above 1,000 rpm, or only below? Step motors have higher magnetic-pole count than servomotors, and the constant-torque range is limited. Above about 1,000 rpm, step motors produce less than 30% of the torque available at low speed. The lower pole count of the servo produces constant torque over a wider speed range — up to about 2,500 rpm.
Size for size, the servomotor produces less torque in lower speed ranges, but at higher speeds the servomotor achieves higher torque output. Figure 1 shows the difference in torque-vs.-speed characteristics. The motors are the same size and use similar amounts of copper, steel, and magnetic materials.
Can the application tolerate position loss? Most step motors operate open loop; there is no feedback on the motor’s true position. If the step motor loses position, you must manually reset it. Servomotors usually operate closed loop; the controller always knows the motor’s position and calls for the right current to maintain the commanded position.
Do you need positioning resolution higher than 1.8 deg? Position accuracy and position resolution are frequently confused. Although higher accuracy arises from higher resolution, higher resolution is not the single determinant of high accuracy. Parameters such as load, friction, and servo controller gain are also important. Typically, step-motor systems have motors designed for full torque production every 1.8 mechanical degrees; that is, for 200 steps/rev. The 200 step/rev is the full-torque resolution of the motor. A servomotor using a typical resolver or encoder for closedloop feedback, usually has at least 1,000 steps/rev as positioning resolution.
Electronic manipulation of step motors, called microstepping, improves resolution significantly, but cannot offer full torque production at resolutions higher than that of a 1.8-deg step motor. Accuracy is different. For example, a microstepping system with resolution of 25,000 steps/rev will lose about 1 arcminute of positioning accuracy for every ounce-inch of torque applied to the motor shaft. Here, a motor that drives a load having 7 oz-in. of friction torque will be only as accurate as ± 7 arc-min. This is about the same as a servomotor with an encoder.
Continue on Page 2
Another means of shortening stepmotor steps is the 5-phase method, which adds a pole pair to the stator of a regular 2-phase hybrid step motor. This changes the rotor-stator offset of a 2- phase step motor from one fourth to one tenth the rotor pitch. The 5-phase motor has much lower irregular rotations compared with a full-step, 2-phase motor, and has nearly no resonance effects. Torque ripple of a 5-phase step motor vs. a regular 2-phase step motor is also much reduced. For more about step-motor selection, see Reference (1).
How much load inertia can you tolerate? Maximum inertia mismatch between load and motor for a step motor should not exceed 10:1. If it does, the load could overpower the motor, resulting in stalling or lost steps during operation. Because step motors usually operate open loop, they can fall out of synchronism if overloaded. If this occurs, you must reset the motor manually to the correct position. Because of its overloading capability and closed-loop configuration, a dc servomotor can control load inertias up to 50 times more than motor inertia if the acceleration requirements are not excessive. These ratios are maximums. The motor and load, when connected, may not make a stable combination.
If your application requires high acceleration- deceleration rates or fast cycles, we recommend that load-to-motor inertia ratio of either type motor be reduced to less than 4:1, and in very high dynamic applications, to 1:1.
Matched inertias reduce overshoot and settling time. The least expensive way to match inertias is with mechanical power transmission. It is costly and inefficient to buy the next larger motor size just to equalize the load and motor inertias.
Remember: Each type of motor has its place. Proper selection is essential.
Motor sizing
Once motor type has been determined, it is just as important to pick the right size. A motor that is too large or too small limits performance.
There are several methods you can use to select the correct size step motor or servomotor. The most common technique is to determine the maximum torque and speed the application requires and select a motor that, based on speed-vs.-torque characteristics, meets or exceeds the application requirements.
This is a good way to quantify statically whether the motor will “do the job.” In some applications this may be the only test you need. However, “the job” may call for more than just supplying torque. This method alone won’t reveal potential dynamic instabilities between motor and load, and it won’t insure adequate power transfer from motor to load and vice versa.
Most manufacturers of step and servomotors now select their products with a computer program. In selecting your motor, it is a good idea to get one of these programs and do some trial-and-error sizing to get a feel for the size you need. After that — especially if you have a dynamic application — you should consider the following methods as a second opinion. You may be surprised by the results, and it could save you some time and money.
Balance the load and motor inertias. In rapid start-and-stop applications, you can significantly improve the cycle times by balancing the load inertia and the motor inertia. If the initial selection is adequate, the motor will have the power needed to move the load in the required time, but other issues including frictional damping and overshoot may limit cycle time. For example, rapid acceleration and deceleration may be needed in 100 msec, but if motor inertia is too small relative to load inertia, it may take a lot of time overshooting and settling to a complete stop. A motor with higher torque will not necessarily solve this problem. The motor may be trying to control the load’s inertia, but its own inertia is too small.
Balancing load and motor inertia is the correct way to optimize dynamic performance. With today’s brushless servomotors and high-energy magnetic materials, the motor inertia on many models has decreased. This has created a greater imbalance between motor and load inertia in many applications. Electronic tuning is ineffective if the imbalance is too large and may result in unstable resonance conditions.
Continue to Page 3
Rather than use a bigger motor and add unnecessary cost to the system, try using mechanical means to balance load and motor inertias. First, eliminate as much mass from the load as possible without jeopardizing the equipment. Then try mechanical speed reduction. For example, a timing belt with a 3:1 ratio reduces the effective load inertia to the motor by a factor of 9 (the square of the speed ratio). A timing belt also adds frictional damping and shock-absorbing material without introducing backlash or lost motion. A timing belt can also provide a way to tuck the motor out of the way.
Likewise, do not be afraid to use a suitable gear drive to reduce load inertia. Inertia reduction again is proportional to the square of the speed ratio. Brushless servomotors won’t mind running at higher speeds.
Balance the load to motor “power rate.” A figure of merit often used in motor selection for rapid incremental motion is called “power rate.” Don’t confuse this with a motor’s “power rating.” Power rate determines the rate at which power is transferred from the motor into the load and vice versa. It is the time rate at which a device can convert power into mechanical motion. Power rating has to do with the motor’s steady-state output power.
Balancing the load power rate and the motor power rate can help minimize cycle times. A motor too large for its intended load will limit the application’s optimum cycle time. (Why accelerate more motor inertia than the load needs?) A motor with insufficient power rate will not accelerate the load optimally. A simple power rate calculation will identify the motor’s ability to control loads in dynamic applications. If the power rate of a motor and load are balanced, the system has optimum power transfer within a given cycle time.
Here is a simple test to determine if the motor power rate and the load power rate are balanced: Divide the square of the motor’s peak torque by its inertia.
P'm=Tpk2 ÷ Jm
where:
P'm = Motor power rate
Tpk2 = Motor peak torque
Jm = Motor inertia
These formulas are derived in Reference (2).
In the optimum case, the inertia of the motor and the load are equal, and the motor power rate is four times greater than the load power rate in order to balance. The factor of four is derived from the need of the motor to accelerate its own inertia plus the load inertia. See Reference (2) for more about the factor of four.
If you really want to zero in on the power rate method, try plotting the power rates of several motors against their steady-state output power rating. Then plot the load requirement, remembering that it will be one fourth the motor power rate. The correct motor will be the one that most closely matches the load.
The technique of matching motor and load inertias and selecting the motor based on a balance between power rates provides information about the anticipated success of the application and provides a basis from which motors can be evaluated equally.
References
1. PTD, “How to Choose a Step Motor: A Real Example,” Nick Johantgen, 3/93, p. 52.
2. PTD, “”Power Rate Simplifies Drive Selection,” Frank Arnold, Jr., 6/88, p. 21.
William E. Seitz is President, Raymond Engineering Inc., Subs. Kaman Corp., Middletown, Conn., a maker of defense systems components and other high-precision devices.